290 



generally the cells are somewhat produced below, so as to 

 be pyriform, and are attached to one another by the base 

 of the short peduncular extremity; sometimes three, 

 sometimes two, originate from the top of another cell. 

 The lateral branches do not usually stand out at a right 

 angle, but take an ascending direction. The habit and 

 mode of growth seem to resemble very closely those of 

 the Crag species H. abstersa, Busk. 



The cells, when crowded together, are rather short and 

 ovate ; they are umbonate, and traversed by very distinct 

 transverse striae. The orifice agrees in all respects with 

 that of the normal H. divaricata. The ovicelligerous cells 

 are not smaller than the rest. Notwithstanding the 

 striking peculiarities of this form, I believe that it is 

 properly referred to H. divaricata, which exhibits many 

 variations in the length of the tubular fibre, and not un- 

 frequently has the cells more or less massed together*. 



The var. /3 is characterized by an exaggerated condition 

 of the keel on the front of the cell. When seen in profile 

 it stands out as a prominent ridge, but is abruptly trun- 

 cated immediately below the orifice. 



H. Patagonica, Busk, appears to be nothing more than 

 a thick-walled and rudely annulated variety of the present 

 species. As represented in the British-Museum Catalogue, 

 it has some affinity with our var. conferta ; but the figure 

 in the 'Crag Polyzoa* makes a near approach to the 

 ordinary H. divaricata. 



In characteristic and well-developed specimens of the 

 latter, branching takes place with much regularity at right 

 angles or nearly so to the main lines of cells, and colonies 

 exhibit a very definite dendritic habit. 



* Dr. Johnston mentions a variety "in which the cells are contiguous" 

 and which occurs on sea-weeds only. 



